Honey XI

[peter]

with the latest cur-dev-Honey , all of these are solved quite easily

How about this one, Mike?

[d]1rb1qrk1/2b2pp1/p3pBn1/3pP1Pp/1ppP4/2P1QN2/PP3P1P/R2BR1K1 w - - 0 21

bm Nh4

http://talkchess.com/forum3/viewtopic.p ... 08#p819008

[MikeB]

with the latest cur-dev-Honey , all of these are solved quite easily

How about this one, Mike?

[d]1rb1qrk1/2b2pp1/p3pBn1/3pP1Pp/1ppP4/2P1QN2/PP3P1P/R2BR1K1 w - - 0 21

bm Nh4

http://talkchess.com/forum3/viewtopic.p ... 08#p819008

negative - it appears to be a depth 50+ moves requiring hundreds of billions of nodes to be searched for Honey to find it ...both Honey and Bluefish have it as the 4th move for white in the pv..Bluefish had a positive score with h2h3 much earlier

info depth 53 seldepth 49 multipv 1 score cp 0 nodes 123714302915 nps 82936054 hashfull 1000 tbhits 0 time 1491683 pv h2h3 g8h7 d1c2 a6a5 e3d2 f8h8 f3h4 h7g8 h4g6 f7g6 d2d1 c8d7 e1e3 a5a4 b2b3 c4b3 a2b3 a4a3 c3b4 b8a8 e3c3 c7b6 a1a2 d7b5 c3f3 a8a7 h3h4 a7a8 g1h1 a8a7 f3c3 a7a8

[peter]

negative - it appears to be a depth 50+ moves requiring hundreds of billions of nodes to be searched for Honey to find it ...both Honey and Bluefish have it as the 4th move for white in the pv..Bluefish had a positive score with h2h3 much earlier

info depth 53 seldepth 49 multipv 1 score cp 0 nodes 123714302915 nps 82936054 hashfull 1000 tbhits 0 time 1491683 pv h2h3 g8h7 d1c2 a6a5 e3d2 f8h8 f3h4 h7g8 h4g6 f7g6 d2d1 c8d7 e1e3 a5a4 b2b3 c4b3 a2b3 a4a3 c3b4 b8a8 e3c3 c7b6 a1a2 d7b5 c3f3 a8a7 h3h4 a7a8 g1h1 a8a7 f3c3 a7a8

Thanks for trying.
I know it to be one of the darkest blind spots of A-B engines compared to NN at all.
Yet I think it to be a matter of pruning still,
After 21.Nh4! Nxh4? 22.Qh3 is one simpliest to be got into tree, but much more difficult probably 21...bxc3?! 22.bxc3! instead of 22.Nxg6? might be the main border for excluding the solution at first ply.
Main line 21...Kh7 isn't the big problem anymore then, if the two other seductions are cleared up.